Gluconeogenesis is a complex process, which is difficult to study in vivo Thus, the first aim of the present study was to compare three recently published methods to measure gluconeogenesis using stable i.e. non-radioactive, naturally occurring isotopes ([U-13C]glucose, [2-13C]glycerol and deuterium oxide). The second aim was to determine if VLBW infants can utilize the gluconeogenic pathway to maintain normal blood sugar concentrations when they receive sugar at a reduced rate (to prevent too high blood sugar concentrations) as part of complete nutrition by vein i.e. sugar, fat and amino acids (building blocks of protein). The third aim was to determine whether fat and amino acids are equally used as substrates for sugar production when sugar delivery is reduced. The hypotheses to be tested were: 1. The three methods used to measure gluconeogenesis provide similar results. 2. VLBW infants receiving sugar at a rate less than normal sugar turnover rate (to prevent too high blood sugar concentrations) as part of complete nutrition by vein, maintain normal blood sugar concentrations mainly via gluconeogenesis. 3. The fat content of nutritional solutions given by vein to premature infants is more important in supporting gluconeogenesis than are the amino acids. To test the first two hypotheses, three comparable groups of VLBW infants (a total of 20 infants) were studied on their 5th day of life. They were born after 27+/-0.2 gestational weeks and weighed 996+/-28 grams. Eight infants were studied by [U-13C]glucose, six by [2-13C]glycerol and six by deuterium oxide. All infants received sugar (at a rate corresponding to half their normal sugar turnover rate) plus fat and amino acids by vein for 8-12 h.